AI-powered digital twin for industrial supply chains

Turn fragmented supply-chain data into an intelligent operating model.

LinePulse connects operational data, relationships, and constraints across the supply network so teams can see what is changing, evaluate scenarios, and make better decisions across planning and execution.

Explore the digital twinJoin the design-partner program
  • Evidence-backed outputs
  • Human approval
  • Read-only by design
  • Built for industrial operations

The industrial supply chain is a dynamic system

Every operational change has consequences across the network.

Supply, demand, inventory, production, logistics, and customer commitments are tightly connected. Yet the information needed to understand those relationships remains distributed across enterprise systems, spreadsheets, documents, and operational communication.

The problem is not a lack of data. It is the lack of a shared, decision-ready representation of the system.

When something changes
  • Demand changes
  • Material availability changes
  • Production-capacity changes
  • Logistics changes
  • Inventory constraints

Data and workflows

  • ERP
  • Planning
  • Warehouse
  • Logistics
  • Supplier
  • Spreadsheets

Fragmented operational picture

  • Demand
  • Inventory
  • Capacity
  • Transport
  • Commitments
  • Cost

Slow or uncertain decisions

  • What is affected?
  • Which option is better?
  • Who needs to act?

The digital twin layer

A connected model of how the supply chain actually behaves.

LinePulse creates a decision-grade digital twin that connects operational entities, relationships, constraints, events, and business consequences.

Introduce a change
SuppliersPurchasecommitmentsLogisticsMaterialsInventoryBills ofmaterialProductionresourcesCustomerdemandFinancialexposureSuppliersPurchasecommitmentsLogisticsMaterialsInventoryBills ofmaterialProductionresourcesCustomerdemandFinancialexposure
How it propagates
  1. Materials
  2. Bills of material
  3. Inventory
  4. Production resources
  5. Customer demand
  6. Financial exposure

Which products and orders depend on this material?

Selected in the model

Materials

Parts and materials, with their sourcing and substitutes.

In the prototype model

Conceptual illustration. The current prototype evaluates delivery changes on uploaded CSV and Excel data. Other change types shown here are roadmap capabilities.

From inputs to decisions

  1. InputsOperational data from ERP, planning, warehouse, logistics and suppliers
  2. Operational modelEntities, relationships, constraints and events, connected
  3. ScenariosChanges propagated through the network, alternatives compared
  4. DecisionsEvidence, assumptions and consequences, for human approval

The progression

Solid steps exist in the prototype. Dashed steps are roadmap capabilities.
  1. 1Connect fragmented dataCSV and Excel today
  2. 2Represent the operational systemCanonical model
  3. 4Simulate scenariosDelivery changes today

How it works

Four capabilities, one operating model.

Each capability below shows what the working prototype does today and what is planned.

CONNECT

Bring together structured operational data and unstructured business information.

In the prototype
CSV and Excel uploads, with column mappings confirmed by a person.
Roadmap
Unstructured documents and connectors to existing systems.

MODEL

Represent relationships, dependencies, constraints, and provenance across the network.

In the prototype
A canonical model of suppliers, materials, orders, inventory, BOMs and production, with each fact traced to its source.
Roadmap
Logistics and wider network relationships.

SIMULATE

Propagate changes through the system and compare possible scenarios.

In the prototype
Deterministic impact calculation for a delivery change, with mitigation scenarios ranked side by side.
Roadmap
Demand, capacity and logistics changes, plus prediction.

DECIDE

Present evidence, assumptions, consequences, and recommended actions for human approval.

In the prototype
A decision view where a person approves or adjusts, recorded in an audit log.
Roadmap
Agent-assisted coordination of approved actions.
Illustrative scenario · synthetic data · working prototype

Illustrative scenario: when one change propagates through the network

A change in an operational input is propagated through the digital twin. Here, one delayed delivery shows what each consequence looks like once the network is connected.

  • Affected materials
  • Inventory coverage
  • Production consequences
  • Customer exposure
  • Financial exposure
  • Scenario alternatives
  • Evidence and assumptions
  • Human approval
ScenarioDelivery change · Subframe bracket
Awaiting human approval
Change
Supplier A SUP-0214Delivery moved Sep 24 → Sep 30
+6 dFACT
Commitment
PO #482911,200 pcs · open qty 1,200
FACT
Material
Subframe bracketP/N 88-1043-A · single-sourced
FACT
Inventory
2.1 days on hand · need 6.0
−3.9 dCALC
Production
3 production orders · Line 3Job #47 · Shift B · Sep 27 06:00
CALC
Customer
SO-77120JIS call-off Sep 28
CALC
Financial
Estimated line-stop costIf no action is taken
€4,300CALC
Penalty
Contract late-delivery penaltyNo rate on file, so none is assumed
UNKNOWN
Scenario alternativesDeterministic calculations
OptionAction costResidual exposure
€900€0
€2,600€0
€0€4,300 + unknown
Human approvalSource 40% from alternate supplier

Static illustration of the kind of output the prototype produces. All names and figures are synthetic. This is not a customer case study.

Built for decisions that need to be understood.

LinePulse is deterministic decision support. Figures come from repeatable calculations over source records, so the same inputs produce the same result.

  • Every output has an evidence trailFigures trace back to the source records behind them.
  • Facts and calculations are separatedEach value is labelled as a source fact or a calculated result.
  • Assumptions are visibleThe assumptions behind a result are listed beside it.
  • Unknown values are not silently filled inA missing input is reported as unknown, never assumed to be zero.
  • Human approval remains part of the workflowA person approves or adjusts each recommendation, and the decision is logged.
  • The current prototype does not write to customer systemsIt works from uploaded data and changes nothing in source systems.

Roadmap

From one high-value workflow to an intelligent supply-chain operating layer.

A modular deployment path, designed to build on existing systems and move toward future agentic orchestration one stage at a time.

  1. FOUNDATION

    Data connection, operational model, provenance, evidence

    In the prototypeData connection is file-based today. System connectors are planned.

  2. INTELLIGENCE

    Prediction, risk identification, impact analysis, scenario evaluation

    Partly in the prototypeImpact analysis and scenario evaluation exist today. Prediction and risk identification are planned.

  3. OPTIMIZATION

    Network, inventory, capacity, and planning optimization

    Roadmap directionNot current product functionality.

  4. ORCHESTRATION

    Agent-assisted workflows and increasingly automated coordination

    Roadmap directionNot current product functionality.

Design partnership

Explore a focused design partnership

We are working with a small number of industrial companies to understand where a digital twin can create the highest operational value. The first step is to identify one decision workflow, understand the available data, and test whether the model produces useful, explainable outputs.

  1. Identify one decision workflow
  2. Understand the available data
  3. Test whether the outputs are useful and explainable

Discuss a design partnership

Tell us about your operation and we’ll be in touch.

Prefer to talk first? Book a 20-minute call